ELIZA (1966): the first chatbot people confided in
ELIZA was a computer program written by Joseph Weizenbaum at the Massachusetts Institute of Technology and described in January 1966. It answered typed messages using simple rules, most famously in the style of a psychotherapist, and people who used it poured feelings into the conversation anyway.
What was ELIZA?
ELIZA "is a program operating within the MAC time-sharing system at MIT," Weizenbaum wrote in his paper in Communications of the ACM, volume 9, number 1. It was written in a language called MAD-SLIP for the IBM 7094 computer.
The program looked for keywords in what a person typed and transformed the sentence according to rules. Those rules lived in a separate "script," so ELIZA could play different roles. The best-known script made ELIZA "respond roughly as would certain psychotherapists (Rogerians)," the paper says, a style of therapy that reflects a patient's words back as questions.
Why was it called ELIZA?
Weizenbaum explained the name in the paper: "Like the Eliza of Pygmalion fame, it can be made to appear even more civilized, the relation of appearance to reality, however, remaining in the domain of the playwright." Its language abilities, he wrote, "may be continually improved by a 'teacher.'"
What did a conversation with ELIZA look like?
The paper prints a sample conversation. It begins:
Men are all alike. IN WHAT WAY? They're always bugging us about something or other. CAN YOU THINK OF A SPECIFIC EXAMPLE? Well, my boyfriend made me come here. YOUR BOYFRIEND MADE YOU COME HERE
ELIZA's lines are in capital letters. Every reply is built from the user's own words or a stock question; the program understands nothing.
“once a particular program is unmasked, once its inner workings are explained in language sufficiently plain to induce understanding, its magic crumbles away”
Why did people confide in it?
The Computer History Museum, in its Chatbots Decoded exhibit, says Weizenbaum "became alarmed at how users, captivated by the software, invested chats with emotion and meaning," "even when they knew ELIZA lacked any understanding."
Weizenbaum had anticipated part of this in 1966. A person in conversation, he wrote, defends the impression of being understood "by attributing to his conversational partner all sorts of background knowledge, insights and reasoning ability." He also wanted to break the spell. "But once a particular program is unmasked, once its inner workings are explained in language sufficiently plain to induce understanding, its magic crumbles away," the paper says.
What did Weizenbaum conclude?
He became one of AI's early critics. The museum quotes him in 1976: "we ought not now to give computers tasks that demand wisdom." He argued, the museum says, that values "sprang only from lived experience."
Why ELIZA belongs in the museum
ELIZA's scripts were meant to be written by others: MIT's Education Research Center published an ELIZA Scriptwriter's Manual in 1968, now in the Computer History Museum's archive. Its lasting lesson is the one Weizenbaum drew from it: people readily treat a program that talks like a person as if it understood them.
FAQ
Who created ELIZA?
Joseph Weizenbaum of the Massachusetts Institute of Technology, who described the program in Communications of the ACM in January 1966.
How did ELIZA work?
It found keywords in what a person typed and rearranged the sentence using rules from a script, such as one that imitated a Rogerian psychotherapist.
Why is it called ELIZA?
After Eliza in Pygmalion, because, like her, the program could be taught to appear more civilized, Weizenbaum wrote.
Sources
- Joseph Weizenbaum, ELIZA: A Computer Program For the Study of Natural Language Communication Between Man and Machine, Communications of the ACM 9(1), pp. 36-45, January 1966 (ACM Digital Library)
- Computer History Museum, Chatbots Decoded: Exploring AI · archived copy
- Paul R. Hayward, ELIZA Scriptwriter's Manual, MIT Education Research Center, 1968 (Computer History Museum archive) · archived copy
Cite this
ELIZA (1966): the first chatbot people confided in, New Mind Museum, November 6, 2026. https://newmindmuseum.com/history/2026/11/eliza-1966-weizenbaum-first-chatbot/
New Mind Museum. (2026, November 6). ELIZA (1966): the first chatbot people confided in. https://newmindmuseum.com/history/2026/11/eliza-1966-weizenbaum-first-chatbot/
“ELIZA (1966): the first chatbot people confided in.” New Mind Museum, 6 Nov. 2026, https://newmindmuseum.com/history/2026/11/eliza-1966-weizenbaum-first-chatbot/.
@misc{new-mind-museum-2026-eliza-1966-the-first-chatbot-people-conf,
author = {{New Mind Museum}},
title = {ELIZA (1966): the first chatbot people confided in},
year = {2026},
url = {https://newmindmuseum.com/history/2026/11/eliza-1966-weizenbaum-first-chatbot/}
}Free to quote and cite with a link to this page.
Keep reading
Sources
- Joseph Weizenbaum, ELIZA: A Computer Program For the Study of Natural Language Communication Between Man and Machine, Communications of the ACM 9(1), pp. 36-45, January 1966 (ACM Digital Library)
- Computer History Museum, Chatbots Decoded: Exploring AI
- Paul R. Hayward, ELIZA Scriptwriter's Manual, MIT Education Research Center, 1968 (Computer History Museum archive)